Usherin defects lead to early-onset retinal dysfunction in zebrafish

Abstract

Mutations in USH2A are the most frequent cause of Usher syndrome and autosomal recessive nonsyndromic retinitis pigmentosa. To unravel the pathogenic mechanisms underlying USH2A-associated retinal degeneration and to evaluate future therapeutic strategies that could potentially halt the progression of this devastating disorder, an animal model is needed. The available Ush2a knock-out mouse model does not mimic the human phenotype, because it presents with only a mild and late-onset retinal degeneration. Using CRISPR/Cas9-technology, we introduced protein-truncating germline lesions into the zebrafish ush2a gene (ush2a^rmc1: c.2337_2342delinsAC; p.Cys780GlnfsTer32 and ush2a^b1245: c.15520_15523delinsTG; p.Ala5174fsTer). Homozygous mutants were viable and displayed no obvious morphological or developmental defects. Immunohistochemical analyses with antibodies recognizing the N- or C-terminal region of the ush2a-encoded protein, usherin, demonstrated complete absence of usherin in photoreceptors of ush2a^rmc1, but presence of the ectodomain of usherin at the periciliary membrane of ush2a^b1245-derived photoreceptors. Furthermore, defects of usherin led to a reduction in localization of USH2 complex members, whirlin and Adgrv1, at the photoreceptor periciliary membrane of both mutants. Significantly elevated levels of apoptotic photoreceptors could be observed in both mutants when kept under constant bright illumination for three days. Electroretinogram (ERG) recordings revealed a significant and similar decrease in both a- and b-wave amplitudes in ush2a^rmc1 as well as ush2a^b1245 larvae as compared to strain- and age-matched wild-type larvae. In conclusion, this study shows that mutant ush2a zebrafish models present with early-onset retinal dysfunction that is exacerbated by light exposure. These models provide a better understanding of the pathophysiology underlying USH2A-associated RP and a unique opportunity to evaluate future therapeutic strategies.

Keywords: Retinal dysfunction; Retinitis pigmentosa; Usher syndrome; Usherin; Zebrafish; ush2a.

Figure 1. Schematic presentation of zebrafish usherin

Motif alignment of human and zebrafish usherin. Both proteins have an identical predicted domain architecture. EGF-lam, laminin-type epidermal growth factor-like domain; LamG, laminin G domain; LamG-like, LamG-like jellyroll fold domain; LamNT, laminin N-terminal domain; FN3, fibronectin type 3 domain.

Figure 2. Usherin localizes at the periciliary membrane and accessory outer segments of adult zebrafish photoreceptor cells

(A) Retinal sections of wild-type adult zebrafish are labeled for usherin (green signal) and the connecting cilium marker centrin (magenta signal). (B) Usherin labeling at the periciliary region of rod photoreceptors is present adjacent to the connecting cilium marked by centrin (indicated by an arrow head). (C) In cones, usherin is also detected at the periciliary region (arrow heads) as well as apical to the connecting cilium marked by centrin labeling (arrows). Nuclei were stained with DAPI (blue signal). (D, E) Electron microscopy images of adult zebrafish retinas show that usherin localizes at the periciliary membrane (arrowheads) as well as in the accessory outer segments (AOS; arrow) of cone photoreceptors. Scale bars in A: 10 μm; B: 3 μm; C: 3 μm; D: 1 μm E: 0.1 μm. CC: connecting cilium, COS: cone outer segment, CP: calyceal processes, ONL: outer nuclear layer, OS: outer segment, ROS: rod outer segment, RPE: retinal pigment epithelium.

Figure 3. ush2a transcript analyses in homozygous ush2a^rmc1 and ush2a^b1245 mutant zebrafish

(A) The locations of the mutations in ush2a^rmc1 and ush2a^b1245 are schematically depicted. (B) RT-PCR analyses of ush2a transcripts derived from homozygous ush2a^rmc1 or (C) homozygous ush2a^b1245 larvae (5 dpf) provided no indications for an effect of the introduced lesions on pre-mRNA splicing. WT: wild-type, Mut: homozygous mutant, mQ: milliQ water.

Figure 4. Localization of usherin in frontal cryosections of wild-type and mutant zebrafish retinas (5 dpf)

Zebrafish-specific anti-usherin-C (A) or anti-usherin-N (B) antibodies are employed (red signal). (A) In ush2a^rmc1 as well as ush2a^b1245 homozygous mutant larvae no signal was detected with the anti-usherin-C antibody, whereas in both TLF and AB* wild-type controls usherin was present adjacent to the connecting cilium marked by centrin (green signal). (B) No signal was detected in the retina of homozygous ush2a^rmc1 mutants using the anti-usherin-N antibody, however, a signal at the photoreceptor periciliary region was observed in the retina of homozygous ush2a^b1245 larvae when this antibody was used. In wild-type controls, the anti-usherin-N antibody also detected usherin at the photoreceptor periciliary region. (n= 35 larvae, from 5 biological replicates with TLF and ush2a^rmc1 larvae, n= 64, from 8 biological replicates with AB* and ush2a^b1245 larvae). Nuclei are stained with DAPI (blue signal). Scale bars: 10 μm.

Figure 5. Ablation of usherin affects the localization of Whrna and Whrnb in the zebrafish retina (5 dpf)

(A) Intensity of Whrna labeling was significantly reduced in both ush2a^rmc1 and ush2a^b1245 mutants as compared to corresponding wild-types (n=35 larvae from 5 biological replicates with ush2a^rmc1 mutant larvae and TLF larvae, and n=54 larvae from 7 biological replicates with ush2a^b1245 mutant larvae and AB* larvae). Intensities of fluorescence using antibodies directed against Whrna (red signal) were quantified and plotted as scatter plots next to the corresponding images. Each dot represents the average grey-value per eye (n=5 ush2a^rmc1 eyes and n=6 TLF larvae eyes, p<0.05 and n=10 ush2a^b1245 and n=9 AB* eyes, p=<0.01, two-tailed unpaired Student’s t-test). (B) Whrnb labeling was reduced in ush2a^rmc1 mutants, but unaltered in ush2a^b1245 mutants compared to corresponding wild-types (n=35 larvae from 5 biological replicates with ush2a^rmc1 mutant larvae and TLF larvae, and n=54 larvae from 7 biological replicates with ush2a^b1245 mutant larvae and AB* larvae). Intensity of Whrnb labeling (red signal) was significantly reduced in ush2a^rmc1 larvae as compared to wild-type (TLF) larvae (n=5 ush2a^rmc1 eyes and n=6 TLF eyes, p<0.01, two-tailed unpaired Student’s t-test). In ush2a^b1245 larvae, the intensity of Whrnb labeling appeared unaltered as compared to wild-type (AB*) larvae (n=27 eyes for both ush2a^b1245 and AB* wild-type; p=0.57, two-tailed unpaired Student’s t-test). Nuclei were stained with DAPI (blue signal) and anti-centrin (green signal) was used as a marker for the connecting cilium. Scale bars: 10 μm, dpf: days post fertilization, a.u.: arbitrary units, * indicates p<0.05, ** indicates p<0.01, and ns: not significant.

Figure 6. Quantification of apoptotic events in the ONL of zebrafish ush2a mutants (8 dpf)

Homozygous ush2a^rmc1 and ush2a^b1245 larvae were exposed either to 300 lux in a (14/10h) day/night rhythm or to continuous light at 3000 lux from 5-8 dpf. Individual dots represent the number of apoptotic cells in the ONL per eye. Strain matched wild-type TLF or AB* zebrafish were used as controls. (A) TLF and ush2a^rmc1 larvae kept at 300 lux with a day/night cycle showed on average 0.2 (± 0.1; 30 eyes) and 0.8 (± 0.2; 28 eyes) apoptotic cells per eye, respectively (p<0.05). When kept under 3000 lux of continuous light, 1.7 (± 2.0; 34 eyes) and 5.1 (± 4.5; 28 eyes) apoptotic cells per eye were observed in TLF and ush2a^rmc1, respectively (p<0.01). (B) AB* and ush2a^b1245 larvae maintained at 300 lux with a day/night cycle showed on average 2.5 (± 1.9; 28 eyes) versus 3.8 (± 2.3; 16 eyes) apototic cells per eye, respectively (n.s.). When kept under 3000 lux of continuous light, on average 4.2 (± 1.7; 23 eyes) versus 24.8 (± 7.9; 18 eyes) apoptotic cells per eye were observed in AB* and ush2a^b1245, respectively (p<0.0001). Each point represents the number of apoptotic cells in the ONL per eye and the standard error of the mean are given as error bars. * indicates p<0.05, ** indicates p<0.01, *** indicates p<0.0001, n.s.: not significant (two-tailed Mann-Whitney test); ONL: outer nuclear layer; dpf: days post fertilization.

Figure 7. ush2a mutants show reduced retinal function

Average ERG traces from a cohort of homozygous ush2a^rmc1 and ush2a^b1245 larvae (5 dpf) with age- and strain-matched wild-type controls are depicted. Averages of the maximum ERG amplitudes are plotted as bar graphs ± standard deviation. (A-D) Normalized b-wave amplitudes recorded in both mutants are significantly reduced compared to ERG-traces from wild-type controls (n=72 TLF and n=75 ush2a^rmc1, from 3 biological replicates, p<0.001, two-tailed unpaired Student’s t-test, and n=20 ush2a^b1245 larvae and n=18 wild-type larvae, from 3 biological replicates, p<0.01, two-tailed unpaired Student’s t-test). (E-H) In both ush2a mutants, the a-wave amplitudes are significantly reduced as compared to strain-matched wild-type controls (n=10 for TLF and ush2a^rmc1, from 2 biological replicates, p<0.05, two-tailed unpaired Student’s t-test, and n=19 for AB* and n=23 for ush2a^b1245, from 2 biological replicates, p<0.001, two-tailed unpaired Student’s t-test). * indicates p<0.05, ** indicates p<0.01 and *** indicates p<0.001.

Figure 8. Schematic model of molecular consequences for other USH2 proteins in ush2a mutants

The ectodomain of wild-type usherin associates with the ectodomain of Adgrv1 (indicated by dashed lines). Whrna and Whrnb bind via a PDZ-PBM-mediated interaction to usherin and Adgrv1, respectively. The ush2a^rmc1 allele results in the complete absence of usherin at the periciliary region. As a consequence, localization of Adgrv1, Whrna and Whrnb at the periciliary membrane is also affected. The ush2a^b1245 allele, however, generates a truncated usherin protein that lacks only the C-terminal 62 amino acids and that is still located at the proper subcellular location in the photoreceptor cell. Because Whrna is mislocalized in this mutant and Whrnb localization is largely unaffected, we propose a model in which Whrna has a higher binding affinity for usherin, and Whrnb a higher binding affinity for Adgrv1. IS: inner segment, OS: outer segment, PBM: PDZ-binding motif.